A submillimetre survey of the kinematics of the Perseus molecular

Mon. Not. R. Astron. Soc. 000, 1–15 (2009)
Printed 3 September 2009
(MN LATEX style file v2.2)
A submillimetre survey of the kinematics of the Perseus molecular
cloud: I. data
Emily
I. Curtis1,2? , John S. Richer1,2 and Jane V. Buckle1,2
1
2
Astrophysics Group, Cavendish Laboratory, J. J. Thomson Avenue, Cambridge, CB3 0HE
Kavli Institute for Cosmology, c/o Institute of Astronomy, University of Cambridge, Madingley Road, Cambridge, CB3 0HA
Accepted 2009 September 2. Received 2009 September 2; in original form 2009 July 15
ABSTRACT
We present submillimetre observations of the J = 3 → 2 rotational transition of 12 CO,
and C18 O across over 600 arcmin2 of the Perseus molecular cloud, undertaken with
HARP, a new array spectrograph on the James Clerk Maxwell Telescope. The data encompass
four regions of the cloud, containing the largest clusters of dust continuum condensations:
NGC1333, IC348, L1448 and L1455. A new procedure to remove striping artefacts from the
raw HARP data is introduced. We compare the maps to those of the dust continuum emission
mapped with SCUBA (Hatchell et al. 2005) and the positions of starless and protostellar
cores (Hatchell et al. 2007a). No straightforward correlation is found between the masses of
each region derived from the HARP CO and SCUBA data, underlining the care that must be
exercised when comparing masses of the same object derived from different tracers. From the
13 CO/C18 O line ratio the relative abundance of the two species ([13 CO]/[C18 O] ∼ 7) and their
opacities (typically τ is 0.02–0.22 and 0.15–1.52 for the C18 O and 13 CO gas respectively)
are calculated. C18 O is optically thin nearly everywhere, increasing in opacity towards starforming cores but not beyond τ18 ∼ 0.9. Assuming the 12 CO gas is optically thick we compute
its excitation temperature, Tex (around 8–30 K), which has little correlation with estimates of
the dust temperature.
13 CO
Key words: submillimetre – stars: formation – ISM: kinematics and dynamics – ISM: individual: Perseus.
1
INTRODUCTION
Recent advances in telescope instrumentation have provided an unprecedented view of the star formation process inside molecular
clouds. Near-infrared imaging from e.g. Spitzer (e.g. Evans et al.
2009) captures the youngest stellar objects, whilst (sub)millimetre
continuum imaging with e.g. SCUBA or MAMBO maps the very
earliest stages of star formation, often before a well-defined central
object is established (e.g. Motte, Andre & Neri 1998; Johnstone
et al. 2000; Hatchell et al. 2005; Enoch et al. 2006; Di Francesco
et al. 2008). Simulations of star formation have also become increasingly sophisticated (e.g. Klessen & Burkert 2000; Bate &
Bonnell 2005; Bate 2009a) with the latest models including turbulence, magnetic fields and radiative transfer (Price & Bate 2008;
Bate 2009b). The true, underlying structure in star-forming clouds,
essential for comparison with models, is hard to disentangle from
continuum images alone (Ballesteros-Paredes & Mac Low 2002;
Smith, Clark & Bonnell 2008b): density enhancements along the
line of sight may superpose and limited spatial resolution and sensitivity inevitably blends nearby objects together. Spectral lines from
?
E-mail: [email protected]
c 2009 RAS
molecular species can provide crucial information about the kinematics of and physical conditions inside molecular clouds. They
may for instance allow the separation of multiple objects, moving
at distinct velocities, along a line of sight (e.g. Kirk, Johnstone &
Tafalla 2007). Additionally, line emission is arguably the best discriminator between different support mechanisms for star-forming
cores, i.e. whether magnetic fields or turbulence dominate the required suppression of star formation in the bulk of clouds.
This paper presents a new survey of the kinematics of star formation in the Perseus molecular cloud. We observed four fields,
totalling approximately 600 arcmin2 in the J = 3 → 2 rotational
transition of 12 CO, 13 CO and C18 O using HARP (Heterodyne Array Receiver Programme), a new array spectrograph operating between 325 and 375 GHz on the James Clerk Maxwell Telescope
(JCMT, see Smith et al. 2003, 2008a; Buckle et al. 2009). These
data represent some of the largest and highest-quality maps of the
higher-J transitions of CO and its isotopologues with good angular
resolution (< 20 arcsec). This paper is the first in a planned series
which will examine the gas structure and core kinematics (Paper
II), molecular outflows (Paper III) and the physical conditions in the
cloud cores (Paper IV). This paper is organized as follows: the remainder of the introduction outlines the survey and its aims before
2
E. I. Curtis, J. S. Richer and J. V. Buckle
Table 1. The transitions of CO and its isotopologues observed.
Molecule
CO
13 CO
C18 O
∆J a
νtrans b
(GHz)
Ttrans c
(K)
ncrit d
(cm−3 )
3→2
3→2
3→2
345.796
330.558
329.331
33.2
31.7
31.6
2.54 × 104
2.22 × 104
2.19 × 104
a
Transition quantum numbers.
Rest frequency.
c Height of the upper J level above ground.
d Estimated critical density, assuming a collision cross-section σ =
10−16 cm−2 and velocity 1 km s−1 so ncrit = A × 1010 cm−2 . The Einstein A
values were taken from the LAMDA database (Schöier et al. 2005).
b
Sections 2 and 3 detail the observations and data reduction procedure including an algorithm to calibrate variable sensitivities across
the HARP array for molecular cloud data. Section 4 presents maps
of the integrated intensity of all of the isotopologues while some
initial results including the cloud opacity and excitation temperature, assuming local thermodynamic equilibrium (LTE) are presented in Section 5. We summarize this work in Section 6.
1.1
The survey
The transitions of CO and its common isotopologues in the
345 GHz atmospheric window trace denser and/or warmer regions
than their lower J relatives – compare the temperature of 5 K and
density of ncrit ∼ 700 cm−3 required to collisionally excite the
12 CO J = 1 → 0 line at 115 GHz to Table 1. These conditions are
similar to those inside star-forming SCUBA cores, allowing us to
probe the motions of bulk gas in their vicinity at matched resolution to SCUBA 850 µm observations. This survey is closely linked
to the JCMT Gould Belt Legacy Survey (GBS, Ward-Thompson
et al. 2007). Our data will eventually be included in that analysis
and we share similar science goals, namely:
(i) Map any high-velocity outflowing gas in the 12 CO J = 3 →
2 line to investigate mass-loss and accretion in a large sample of
sources. The presence of an outflow can differentiate between a
protostellar or starless continuum core, establishing the source age.
(ii) Establish the level of non-thermal support inside starforming cores from the C18 O transition, which may also be used
to evaluate the levels of CO depletion.
(iii) Investigate the turbulent structure of the molecular gas.
Perseus (see Fig. 1) is an intermediate star-forming environment between low-mass, quiescent Taurus and high-mass, turbulent Orion (e.g. Ladd, Lada & Myers 1993). The complex consists
of a series of dark clouds at approximately 3h 30m , +31◦ with an
angular extent of about (1.5 × 5) deg, totalling ∼ 1.7 × 104 M
(Bachiller & Cernicharo 1986a). The cloud is associated with
Per OB2, the second closest OB association to the Sun, with an
age < 15 Myr (see Bally et al. 2008). The association has blown
a 20 deg diameter shell of atomic hydrogen into the interstellar
medium with the Perseus molecular cloud embedded in its western
rim. The region is well-studied both as a whole and in individual
sub-sections with a wealth of ancillary data available, particularly
from the COMPLETE project1 (Ridge et al. 2006b).
Notably, large-area surveys of the dust continuum emission
in Perseus have been completed with SCUBA at 850 µm (Hatchell
et al. 2005; Kirk, Johnstone & Di Francesco 2006) and Bolocam at
1.1 mm (Enoch et al. 2006). The 3 deg2 SCUBA survey (Hatchell
et al. 2005) identified 92 submillimetre cores above their completeness limit of 0.4 M (in a 14 arcsec beam). 80 per cent of
these cores were grouped in six clusters: NGC1333, IC348, L1448,
L1455, B1 and B5. The larger and more sensitive Bolocam survey (7.5 deg2 down to 0.18 M ; Enoch et al. 2006) identified 122
compact cores. However, only 5–10 cores were in areas not observed with SCUBA by Hatchell et al., illustrating that most of
Perseus is devoid of active star formation. In this survey we target
the clusters of star-forming cores (see Section 2). Spitzer has provided a census of deeply embedded young stellar objects (YSOs)
in many nearby molecular clouds (e.g. Evans et al. 2009) and this
population has been associated with continuum cores identified by
SCUBA (Jørgensen et al. 2007a, 2008) or Bolocam (Enoch et al.
2008, 2009). In this survey, we use the catalogue of Hatchell et al.
(2007a), who identified the SCUBA cores from Hatchell et al.
(2005) as starless, Class 0 or Class I protostars on the basis of their
spectral energy distributions (SEDs), which incorporated Spitzer
fluxes from the IRAC wavebands.2
Estimates of the distance to Perseus vary from 220 pc (Černis
1990) to 350 pc (Herbig & Jones 1983). Larger distances are often
based on the Perseus OB2 association which has an established distance of ∼ 320 pc from Hipparcos (de Zeeuw et al. 1999). However,
there is some evidence that Per OB2 may lie behind the molecular
clouds of interest which are probably at closer to 250 pc (Černis
1993). In fact, it may not be appropriate to use a single distance
for the entire cloud and extinction studies suggest increasing distances from 220 to 260 pc from west to east (previous references
and Černis & Straižys 2003). Furthermore many authors have suggested that Perseus is a superposition of a least two smaller clouds
– the closer is thought to be an extension of Taurus with the more
distant a shell-like structure (e.g. Ridge et al. 2006a). The latest and
possibly most reliable measure of the distance to the H2 O maser in
NGC1333 SVS13 found (235 ± 18) pc using its parallax at radio
frequencies (Hirota et al. 2008). In this survey we assume Perseus
to be a single entity at a distance of 250 pc for consistency with the
majority of recent studies e.g. the Spitzer Cores to Disks team, c2d
(Evans et al. 2003, 2009) and the Bolocam Perseus survey (Enoch
et al. 2006).
2
OBSERVATIONS
We selected the four largest clusters of continuum cores –
NGC1333, IC348, L1448 and L1455 – to map in the 12 CO, 13 CO
and C18 O J = 3 → 2 lines (see Fig. 1; the exact areas are detailed in
Table 2). All the regions were observed in the three tracers except
for NGC1333 where only 13 CO and C18 O data were taken since
it had already been observed in 12 CO J = 3 → 2 with HARP (J.
Swift, personal communication).
The data were taken as part of the Guaranteed Time programme for the HARP instrument team, over a period of nine nights
between 17th December 2007 and 12th January 2008 with one map
of NGC1333 taken on 28th July 2007. The data comprise ∼43 hrs
total observing time, approximately 25 hrs on sky translating to an
1
the COordinated Molecular Probe Line Extinction Thermal Emission
survey of star-forming regions, see http://www.cfa.harvard.edu/
COMPLETE
2
IRAC (Infrared Array Camera) on Spitzer has four channels at 3.6, 4.5,
5.8, and 8 µm.
c 2009 RAS, MNRAS 000, 1–15
The kinematics of the Perseus molecular cloud
1.5
2
2.5
3
3.5
31:00
30:00
30
Declination (J2000)
30
32:00
1
3
45
3:40
35
30
25
Right ascension (J2000)
Figure 1. Overview of the molecular gas in the Perseus molecular cloud. The grey-scale is C18 O J = 1 → 0 integrated intensity in K km s−1 from the FCRAO
14-m telescope (Hatchell et al. 2005). The dotted contour encloses regions where the visual extinction is > 3.5, from the map of the COMPLETE team (Ridge
et al. 2006b). The four fields observed in this survey are enclosed in boxes: NGC1333, IC348, L1448 and L1455. Also marked is the B1 molecular ridge. B5
the other large cluster of SCUBA cores lies roughly 45 arcmin to the northeast.
Table 2. The regions observed.
Region
Centre a
RA
(h m s)
NGC1333
IC348
L1448
L1455
a
03:28:56
03:44:14
03:25:30
03:27:27
Dec
(◦
0 00 )
+31:17:30
+31:49:49
+30:43:45
+30:14:30
Dimensions
Width
Height
(arcsec)
(arcsec)
760
1100
800
880
900
650
500
520
Area
(arcmin2 )
190
199
111
127
Position of the map centre in J2000 coordinates.
observing efficiency of 58 per cent. The weather conditions were
good to excellent throughout with median receiver and system temperatures of Trx = 180 K and Tsys = 325 K for 12 CO compared to
Trx = 179 K and Tsys = 415 K for 13 CO/C18 O.
The HARP imaging array comprises 16 SIS detectors, arranged on a 4 × 4 grid, separated by 30 arcsec. This under-samples
the focal plane with respect to the Nyquist criterion (λ /2D is 6.0
and 6.2 arcsec at 345 and 330 GHz respectively), therefore standard
HARP observing strategies take data points in between the nominal detector positions to produce a fully-sampled map (see Buckle
et al. 2009). The observations for this survey were taken in the
raster mode at the default sample spacing of 7.3 arcsec. The 13 CO
and C18 O observations were taken simultaneously, utilizing the capability of the new back-end correlator, ACSIS (Auto-Correlation
Spectral Imaging System, Buckle et al. 2009), to split its bandpass
into two separate sub-bands. Both sub-bands provide ∼250 MHz
of bandwidth on the two lines with a channel spacing of 61 kHz
corresponding to approximately 0.05 km s−1 . The 12 CO observations were also taken using a sub-band mode, with both centred at
the same frequency: (i) for high-velocity gas with a bandwidth of
c 2009 RAS, MNRAS 000, 1–15
∼1 GHz, broken into channels 977 kHz wide (0.8 km s−1 ) and (ii)
to trace the gas in detail with ∼250 MHz bandwidth at a channel
spacing of 61 kHz (0.05 km s−1 ).
2.1
Scan strategy
During raster or “on-the-fly” mapping, the telescope scans continuously along a set direction (usually parallel to the longest side
of the map), dumping the data gathered at discrete intervals – as
little as 100 ms apart with HARP/ACSIS. Once the telescope has
scanned the length of the map, it steps perpendicular to the scan
direction and begins a new row. The HARP array is angled at
arctan(1/4) ∼ 14 deg with respect to the scan direction to produce
a fully-sampled image on a 7.3 arcsec grid (Buckle et al. 2009).
The detectors in HARP have varying responses across the array. Additionally, in any one observation the data from a number of
detectors can be unusable with poor baselines. These in combination can lead to maps with variable noise or entirely blank strips.
Our observing strategy employed three factors to reduce these effects:
(i) The perpendicular spacing between scan rows was set to be
half the width of the tilted array. Thus, in any row, half of the detectors are going over previously observed positions and half are
scanning new parts of the sky.
(ii) The most effective way to ‘fill-in’ entirely missing rows is
to ‘basket-weave’ i.e. re-observe the map with the scan direction
perpendicular to the original.
(iii) A small offset between map repeats, of one pixel’s length
perpendicular to the scan direction, will cause the new spectra from
one detector to coincide with those from a different one in the previous repeat.
Each area was observed at least four times, with one repeat offset
4
E. I. Curtis, J. S. Richer and J. V. Buckle
Table 3. Reference positions used.
RA (J2000)
(h m s)
Dec (J2000)
(◦ 0 00 )
Sub-Region used
03:29:00
03:48:38
03:33:11
+31:52:30
+31:49:39
+31:52:02
NGC1333
IC348
L1448 and L1455
lowed. Calibration spectra were taken frequently towards CRL 618
of the 12 CO or 13 CO J = 3 → 2 lines as appropriate. The intensity
in the reference detector was then compared to previously recorded
standards and only if they matched within a calibration tolerance,
were any subsequent observations allowed to continue. All the data
products and images we present are on the antenna temperature
scale (TA∗ , Kutner & Ulich 1981), which can be converted to main
beam brightness temperature, Tmb , using Tmb = TA∗ /ηmb . The efficiency we use, ηmb = 0.66, was measured during the commissioning of HARP.
3
DATA REDUCTION
We used the Starlink software collection3 for the analysis and reduction of HARP data. First, bad or extremely noisy data were
flagged in the supplied HARP time-series format. The resultant
spectra were placed on to a spatial grid using the SMURF reduction package (Jenness et al. 2008). Our final data products are sampled on a 3 arcsec grid, using a 9 arcsec full-width half-maximum
(FWHM) Gaussian gridding kernel, resulting in an equivalent
FWHM beam size of 17.7 and 16.8 arcsec for 13 CO/C18 O and
12 CO respectively. After gridding, a linear baseline was removed
from each spectrum by fitting line-free portions using the Starlink
KAPPA applications. The spectra were also re-binned spectrally to
resolutions of 0.15 km s−1 for 13 CO/C18 O and 1 km s−1 for the
low-resolution 12 CO data.
Figure 2. Representative maps of the RMS noise in K towards IC348 in the
three CO isotopologues. Top: 12 CO J = 3 → 2 noise in 1 km s−1 channels.
Middle and bottom: 13 CO and C18 O noise in 0.15 km s−1 channels respectively.
from the original and the other two basket-weaved variants of the
original and offset scans. Maps of the resulting root mean square
(RMS) noise for the three transitions towards IC348 are shown in
Fig. 2.
2.2
References and calibrations
Finding reference positions has proved considerably harder for
HARP than single-pixel receivers, as all 16 detectors must point
at line-free regions, requiring a ‘blank’ piece of sky approximately
2 arcmin square. Three such positions were used for our Perseus
survey (see Table 3). Separate one minute ‘stare’ observations were
undertaken at 12 CO J = 3 → 2 towards these positions, using references even further from the cloud. All were found to have no
emission above the noise, re-binned in 1 km s−1 velocity channels,
in every working detector.
All the standard telescope observing procedures were fol-
3.1
The HARP flatfield
After the basic data reduction, distinctive stripes were apparent in
the integrated intensity images. Near the peak of the line intensity,
rows of high and low pixel values lay parallel to the scan direction,
implying certain detectors were systematically higher or lower by
as much as a factor of two. The worst affected data were from the
13 CO and high-resolution 12 CO scans (see Fig. 3). A similar pattern
was clear in every individual scan.
The origin of such systematic calibration differences between
detectors is unknown but is likely to be in the intermediate frequency system. However, for the purposes of this paper we require
only a pragmatic method to eliminate the striping artefacts from
molecular cloud data. Therefore, we derive a set of temperature
conversion factors (TCFs), by which we can multiply the spectra
from each detector to get them on to a common intensity scale. This
procedure is similar to the ‘flatfield’ procedure which accounted
for the flux conversion factors with SCUBA (Holland et al. 1999),
therefore we refer to the analogous flatfield for HARP.
Ideally, to derive these factors from observations, one would
3
Now maintained and developed by the Joint Astronomy Centre, see
http://starlink.jach.hawaii.edu
c 2009 RAS, MNRAS 000, 1–15
31:56
58
32:00
Declination
02
04
06
08
The kinematics of the Perseus molecular cloud
50
40
30
20
10
Right ascension
3:44:00
50
40
43:30
45:00
50
40
30
20
10
Right ascension
3:44:00
50
40
43:30
31:56
58
32:00
Declination
02
04
06
08
45:00
ic348_13co_col.sdf
3:44:14.510 32:01:35.16 J2000
08
Starlink GAIA::Skycat
IC348
Jul 02, 2008 at 21:10:23
32:00
observe a small source, providing a fully-sampled map for each detector, so that the total flux can be compared between them. However, we must derive the factors from the scientific observations
themselves, comparing the intensity detector-by-detector for an entire map. First, datacubes were produced for each detector individually and the spectra integrated over the line and summed over the
map. We divided the sum from the reference detector (R11, one of
the central four in the array, see Buckle et al. 2009 for an explanation of how the JCMT points and tracks with HARP) by those
from each of the other working detectors in turn (up to 15 in number), to calculate the respective TCF. The reference detector was
taken as the intensity standard as it is not masked out in any of our
observations. Additionally, all the calibrations are centred on it, so
we have a robust determination of its performance. Once the factors are known, the output from each detector is multiplied by its
respective TCF. The procedure worked very effectively for most of
the data as shown in Fig. 4 and all the subsequent images in this
paper have been multiplied by TCFs, except the C18 O data towards
L1455 which had an insufficient signal-to-noise ratio (SNR)and all
the low-resolution 12 CO datasets, which did not display the artefacts.
It is perhaps surprising that this procedure worked so effectively and undoubtedly the nature of the targets contributed to its
success. The method depends on each detector receiving exactly the
same amount of emission. Offset map repeats and basket-weaving
helped each detector to sample a larger proportion of the target
field. However, it is the uniformity of a molecular cloud’s emission
that means, in general, there is little difference in intensity between
the different detectors. If there were many compact sources or intense emission towards the edges of these maps, which are only
observed by a few detectors, then this technique would be ineffective. Other datasets taken in the same period show similar striping
artefacts, notably those of the JCMT GBS, who have implemented
our algorithm across a number of their fields.
31:52
54
56
58
Declination
02
04
06
eic22
5
45:00
50
40
30
20
10
3:44:00
50
40
43:30
4
RESULTS
Right ascension
NGC1333
32:00
58
31:52
54
56
Declination
02
04
06
08
4.1
45:00
50
40
30
20
10
3:44:00
50
40
43:30
Right ascension
Figure 3. Integrated intensity images for the different J = 3 → 2 transitions
towards IC348 displaying the striping artefacts from our basic data reduction. The spectra are distributed on a 7.3 arcsec grid using nearest-neighbour
allocation to emphasize detector based variations. From top to bottom are
C18 O, 13 CO, 12 CO low-resolution and 12 CO high-resolution maps. The
12 CO data from a few telescope scan rows appear to have slipped, this is
a known problem where the telescope does not return to the top of the map
before commencing integrating.
c 2009 RAS, MNRAS 000, 1–15
NGC1333 is a young stellar cluster in the west of Perseus, associated with the reflection nebula of the same name and the dark
cloud L1450. It has been widely studied (see Walawender et al.
2008 for a review) and is the most active region of star formation
in the Perseus complex. The stellar cluster is very young (< 1 Myr)
and highly obscured, containing about 150 stars totalling 79 M
(Lada, Alves & Lada 1996; Wilking et al. 2004). The latest infrared (IR) survey with Spitzer (Jørgensen et al. 2006; Rebull et al.
2007; Gutermuth et al. 2008) identified 137 objects in the cluster:
98 pre-main sequence stars and 39 protostars. These protostars are
correlated with the position of dense molecular material and dust.
In total, there is approximately 450 M of gas in the region (Warin
et al. 1996). Much of this gas lies in dense filaments, surrounding
cavities (Lefloch et al. 1998; Quillen et al. 2005). The YSOs in
NGC1333 also drive a large number of overlapping outflows, providing one of the clearest examples of the self-regulation of star
formation (Knee & Sandell 2000).
We present maps of the 13 CO and C18 O HARP data in Fig.
5, alongside the SCUBA 850 µm emission over the observed region, originally from observations by Sandell & Knee (2001) but
later analysed as part of the survey of Hatchell et al. (2005). Qualitatively, the C18 O and SCUBA maps closely resemble each other.
E. I. Curtis, J. S. Richer and J. V. Buckle
31:20
10
10
12
12
14
14
16
18
Declination
18
16
Declination
31:20
22
22
24
24
6
30
20
10
3:29:00
50
40
30
28:20
30
20
10
Right ascension
50
40
30
28:20
40
30
28:20
26
26
24
24
22
22
18
14
16
Declination
31:20
31:20
18
12
10
30
20
10
3:29:00
50
Right ascension
40
30
28:20
08
Declination
16
14
12
10
08
3:29:00
Right ascension
30
20
10
3:29:00
50
Right ascension
Figure 4. Action of the flatfield procedure on 13 CO integrated intensity images towards NGC1333. The panels are from top to bottom and left to right: data
from a single scan, unaltered from the basic reduction; the same data as in the previous panel with the flatfield applied; data from all the NGC1333 scans,
unaltered from the basic reduction; and the same data as in the previous panel but with the flatfield applied.
The central cavity, just north of SVS13 is clear in all the tracers,
as is the horseshoe of emission enclosing its northern boundary.
This would suggest that similar material is being explored in both
the C18 O and SCUBA data. However, there are clear differences
in the structures as well. IRAS4 and IRAS2 (Jennings et al. 1987),
are two of the brightest sources in the SCUBA field, yet they have
much weaker C18 O emission and can scarcely be discerned in the
13 CO data. The string of cores to the north of the central cavity,
(sources 54, 56 and 664 ) have quite weak SCUBA emission but are
the brightest objects in the C18 O maps.
4.2
The IC348 molecular ridge
Our second field is in the vicinity of another young IR cluster
(∼2 Myr old), IC348 (see Herbst 2008 for a review), in the east
of Perseus. It comprises several hundred members totalling about
160 M (Luhman et al. 2003). The low stellar disc fraction and lack
of outflow activity may indicate that the cluster is coming to the end
of its star-forming phase (Luhman et al. 1998). In fact the region
we refer to as IC348 hereafter (Fig. 6) is really a bright molecular
ridge some 10 arcmin southwest of the cluster, sometimes associated with the “Flying Ghost Nebula” (Boulard et al. 1995). This
area, in contrast to IC348, is currently undergoing star formation
4
The source numberings we refer to in the rest of this paper are the catalogue number from Hatchell et al. (2007a)
with many embedded objects and outflows, possibly triggered by
the nearby cluster (e.g. Bally et al. 2008). The best-known feature
in the area is HH211, a highly-collimated bipolar outflow driven
by a Class 0 protostar discovered by McCaughrean, Rayner & Zinnecker (1994), which has been the target of many interferometric
studies subsequently (e.g. Gueth & Guilloteau 1999; Chandler &
Richer 2001).
In our C18 O data, depicted in Fig. 6, the close correspondence
with the SCUBA emission is striking. Again there is not a simple
scaling between SCUBA flux and C18 O integrated intensity, for
example the brightest C18 O core is the starless source 17, while the
driving source of HH211 (source 12), the brightest SCUBA core, is
less prominent in C18 O. The 12 CO data looks very different, as it is
probably optically thick (thus tracing only outer layers of the cloud)
and has very bright outflow lobes. Two bipolar outflow structures
can be seen in the 12 CO integrated intensity and are more obvious
in the false colour red-green-blue (RGB) image of the same gas.
The first highly symmetric outflow (driven by source 12) is HH211,
while the second is the more confused north-south flow from source
13 – known also as IC348-MMS (Eislöffel et al. 2003) or IC348SMM2 (Tafalla, Kumar & Bachiller 2006; Walawender et al. 2006).
The bipolar outflow discovered by Tafalla et al. (2006) from IC348SMM3 (source 15), slightly west of IC348-MMS, is also faintly
discernible. Two Class I protostars (sources 14 and 101) are very
red in the RGB image. This may indicate either an outflow where
only one lobe is visible due to e.g. an inhomogeneous environment
c 2009 RAS, MNRAS 000, 1–15
7
The kinematics of the Perseus molecular cloud
1000
1500
2000
2500
1
3000
2
3
4
5
6
8
7
9
10
26
26
500
24
24
57
67
63
31:20
Declination (J2000)
68
60
46
18
Bolo44
55
62
50 52
44
14
48
14
51
16
43
16
Declination (J2000)
45
53
31:20
70
22
47
54 56
18
22
66
49
41
59 42
12
12
65
74
10
10
72
30
20
10
50
3:29:00
40
28:30
30
20
10
5
10
20
10
Right ascension (J2000)
2
3
4
5
6
9
10
15
50
20
40
25
28:30
30
24
26
22
24
Declination (J2000)
16
18
31:20
22
31:20
18
10
10
12
12
14
14
16
Declination (J2000)
3:29:00
Right ascension (J2000)
8
7
26
1
30
20
10
3:29:00
50
Right ascension (J2000)
40
28:30
30
3:29:00
50
40
28:30
Right ascension (J2000)
Figure 5. Overview of the data towards NGC1333. Top left: SCUBA 850 µm emission with contours at 100, 200, 400, 800, 1600 and 3200 mJy beam−1
(Hatchell et al. 2005) with a colour-scale as shown in mJy beam−1 . The peak of submillimetre cores identified and classified by Hatchell et al. (2007a)
are labelled with their Hatchell et al. source number and marked with: black and white circle (starless cores), light blue diamonds (Class 0 protostars) or
purple triangles (Class I protostars). Sandell & Knee (2001) originally presented this data and located many submillimetre sources. Additionally labelled are
prominent sources in the field, NGC1333-IRAS2A (Jennings et al. 1987; Blake et al. 1995; Looney, Mundy & Welch 2000), NGC1333-IRAS4A to C (Jennings
et al. 1987; Lay, Carlstrom & Hills 1995; Rodrı́guez, Anglada & Curiel 1999; Looney et al. 2000), SVS13 (Strom, Vrba & Strom 1976; Haschick et al. 1980;
Snell & Edwards 1981; Grossman et al. 1987; Hirota et al. 2008), HH7-11 (Herbig 1974) and NGC1333-IRAS7 (Lightfoot & Glencross 1986; Jennings et al.
1987; Cohen, Jones & Hereld 1991; Jørgensen et al. 2007b). Top right: Contours of SCUBA 850 µm emission (as previously) overlaid on the C18 O J = 3 → 2
R
integrated intensity as displayed on the bottom left. Bottom left: Integrated C18 O J = 3 → 2 intensity, TA∗ dv, from 5 to 10 km s−1 with contours from 1 to
R ∗
−1
−1
13
10 K km s in steps of 1 K km s . Bottom right: Integrated CO J = 3 → 2 intensity, TA dv, from 2 to 17 km s−1 with contours from 3 to 33 K km s−1 in
steps of 3 K km s−1 .
c 2009 RAS, MNRAS 000, 1–15
8
E. I. Curtis, J. S. Richer and J. V. Buckle
or that the 12 CO emission comes from the protostars themselves
and they are moving with respect to the ambient gas.
4.3
L1448
The dark Lynds cloud L1448 (see Fig. 7), the most westerly of
our targets, is a region dominated by outflow activity. Half a
dozen or so YSOs reside in its dual-core molecular structure of
∼100 M (Bachiller & Cernicharo 1986b; Wolf-Chase, Barsony
& O’Linger 2000). Its outflows are well studied, particularly the
highly-collimated, symmetric flow originating from L1448-C, one
of the youngest known at the time (e.g Bachiller et al. 1990; Bally,
Lada & Lane 1993). Given that the energy in the outflows exceeds
the gravitational binding energy of the cloud, L1448 is likely to be
dispersed by its outflow activity (Wolf-Chase et al. 2000).
The HARP C18 O data seems to reflect straightforwardly the
SCUBA emission in Fig. 7, with the brightest cores at 850 µm also
having the greatest C18 O integrated intensity. Most of the SCUBA
cores also appear as clumps of C18 O emission. The same structure
is clear in 13 CO as well, even though there is some hint that the
emission is lined up along the main southeast-northwest outflow
axis. The outflows themselves dominate the 12 CO data with perhaps three to four flows overlapping. L1448C (source 29) drives
a highly collimated outflow discovered by Bachiller et al. (1990),
which was one of the highest-velocity and youngest found at the
time (Bachiller et al. 1990, 1995). Its northern, blue-shifted lobe intersects an outflow from a cluster of Class 0 objects: L1448N:A/B
and L1448NW. Finally, west of this cluster there are perhaps two
outflows emanating from the vicinity of L1448-IRS2 (sources 30
and 31).
4.4
L1455
Our final field is towards another Lynds cloud, L1455 (Fig. 8) in
the south-west of the complex with a mass of some 40–50 M
(Bachiller & Cernicharo 1986a). Although the smallest and faintest
of our targets, it has some of the most interesting outflow structure.
The C18 O emission is weak and only detectable towards the handful of protostellar and starless cores in the southeast of the field.
The 13 CO data also mainly pick out the compact sources with associated SCUBA emission. One of the brightest 13 CO clumps (also
detected in C18 O at 3h 27m 33s , 30◦ 120 4500 ) has no compact SCUBA
emission and appears coincident with collimated blue-shifted emission from RNO15-FIR (L1455-FIR, source 35), a Class I source associated with the red reflection nebula of the same name. The 12 CO
maps are more intriguing, with a prominent northwest-southeast
CO outflow (Goldsmith et al. 1984; Levreault 1988) some distance
from the cluster of protostars. Either this is driven by an unknown
low-luminosity source or one of the known protostars. The latter
explanation is promoted by Davis et al. (2008) although the structure of the outflow might suggest the former is more likely.
5
5.1
DISCUSSION
Average region properties
In Table 4, we list the masses of the regions derived from the CO
isotopologues and SCUBA 850 µm data. For the continuum data,
we follow Visser, Richer & Chandler (2001) in calculating the region masses, assuming a constant dust temperature and opacity of
12 K and κ850 = 0.012 cm2 g−1 respectively. The gas masses assume each spectral line is optically thin, in LTE and its emitting
gas has an excitation temperature, Tex = 12 K. We use a standard
method to derive the molecular column densities (e.g. Minchin,
White & Padman 1993), which for a CO isotopologue, X, yields:
Z
Ttrans
Tmb dv cm−2 ,
(1)
N(X) = 5 × 1012 Tex exp
Tex
where Ttrans is 33.7, 31.8 and 31.6 K for the J = 3 → 2 transition of
12 CO, 13 CO and C18 O respectively and the integral is in K km s−1 .
To calculate total masses from the derived column densities we assume molecular abundances of 10−7 , 10−6 and 10−4 for C18 O,
13 CO and 12 CO respectively relative to neutral hydrogen. A mainbeam efficiency of ηmb = 0.66 was used for all the isotopologues
as measured during HARP’s commissioning. In each map, only positions with peak values greater than three times the average noise,
hσRMS i, have been included.
The pattern of the different masses is intriguing and not
purely what is expected given the different optical depths of
the CO isotopologues. NGC1333 clearly shows the naivelyexpected trend; as the CO lines become increasingly optically thick
(C18 O→13 CO→12 CO), only the cloud surface layers are probed,
the total emission falls and the mass decreases. The SCUBA
850 µm emission, although optically thin almost everywhere, is
less sensitive than the CO observations to large-scale structure
(& 2 arcmin), due to the spatial chopping in the observing technique. Therefore, it potentially misses much of the cloud mass.
This simple prescription is not followed in any of the other regions, except perhaps L1455. However, here the C18 O emission
is particularly weak so most of the spectra have peaks < 3σRMS
and are therefore not included in the mass estimates of Table 4,
which probably relate to three or four cores. This is a potential reason why the C18 O mass is so much smaller than that from SCUBA.
IC348 and L1448 make an interesting contrast: the C18 O and 13 CO
masses are not very different, with the 13 CO mass marginally larger
in both cases, implying the 13 CO gas is probably optically thin and
more widespread than the C18 O. However, L1448 has a SCUBA
mass some 3–4 times larger than these gas estimates in contrast
to IC348 whose SCUBA mass is 2–3 times smaller. C18 O excitation effects may explain the large C18 O mass in IC348, if there is
a significant amount of subthermal excitation, the C18 O emission
for a given quantity of gas would be boosted relative to LTE causing an over-estimate of the C18 O mass. In L1448, the small C18 O
mass is unlikely to be caused by subthermal excitation and probably arises from a different gas-to-dust ratio or if we underestimate
the dust temperature (which in any case will vary from our constant
assumption of 12 K). In Section 5.3, we show the dust temperature
in L1448 is always higher than the excitation temperature of the
gas. If we underestimate the dust temperature then we overestimate
the corresponding dust mass, which may account for some of the
discrepancy between the tracers.
Average spectra in the different regions are plotted in Fig. 9.
All three isotopologues are centred at similar velocities, approximately 7.8, 8.8, 4.3 and 5.0 km s−1 for NGC1333, IC348, L1448
and L1455 respectively. The strength of the lines are in order of
their abundance as we would expect and these averages show no
global evidence for multiple components or self-absorption. However, individual 12 CO spectra do show self-absorption and/or multiple components. In L1448, the 13 CO line is nearly the same
strength as the 12 CO, implying the 12 CO line is either saturated
(and therefore the gas has a low physical temperature) or there are
significant optical depths in the 12 CO gas due to density or temc 2009 RAS, MNRAS 000, 1–15
The kinematics of the Perseus molecular cloud
Table 4. Region masses from the various HARP J = 3 → 2 integrated intensities and SCUBA 850 µm emission. All spectral-line estimates assume
LTE, τ 1 and Tex = 12 K, while the SCUBA estimate assumes a constant
dust temperature of 12 K. Only points with peaks > 3hσRMS i have been
included.
Region
850 µm
NGC1333
IC348
L1448
L1455
250
71
153
45
Masses (M )
13 CO
C18 O
412
160
38
7
301
212
45
35
12 CO
13
35
9
12
perature effects. Finally, even an average 12 CO spectrum contains
significant high-velocity linewings and there are weaker linewings
clear in the 13 CO spectra towards L1448 and IC348.
C18 O is an ideal tracer of the bulk motions of the molecular
gas, given that it is nearly optically thin throughout and detectable
over larger areas than denser-gas tracers. We plot C18 O line centres
across the regions in Fig. 10. The centres are from non-linear least
squares fitting of Gaussian functions to every spectrum with a peak
> 3σRMS (the noise calculated on each individual spectrum), performed using software from S. Graves. There are many plausible
explanations for variations in velocity across a cloud region:
(i) Rotation. A systematic shift in the line velocity along a particular direction is often suggested to be a signature of bulk rotation.
For instance the Serpens cloud core is thought to be undergoing a
global east to west rotation (e.g. Olmi & Testi 2002) which manifests itself in a very ordered change in the C18 O J = 3 → 2 velocity
observed with HARP by the GBS (Graves et al., in preparation).
In the Perseus regions there are few such global gradients. Ho &
Barrett (1980) inferred a south to north rotation of NGC1333 from
NH3 observations, although Walsh et al. (2007) (and this work) find
no such global gradient in N2 H+ , suggesting Ho & Barrett were biased by the red-shifted gas around IRAS7 (in the centre-left of our
map).
(ii) Outflows. On large scales in our fields, CO outflows do not
seem to correlate with the C18 O centres (although there is some link
on the scale of individual cores; Curtis & Richer, in preparation).
(iii) Constituent clumps moving with respect to the bulk cloud.
Example SCUBA cores seem to enclose regions of very different
velocity, which implies some of the variations trace clumps of gas
that move at distinct velocities with respect to the ambient cloud.
(iv) Gas Flows. In a similar fashion to (iii) the variations in velocity could reflect larger flows of gas. In gravoturbulent models we
expect cores to form at stagnation points, where convergent streams
of material collide (e.g. Padoan et al. 2001). In these data, there are
large areas of coherent motion but the cores do not seem to occur
solely where they intersect.
5.2
13 CO/C18 O
ratio
The optical depth of a spectral line provides important evidence as
to where in the cloud it has been emitted. Optically-thin molecular
lines can provide a trustworthy measure of the column density and
conditions throughout the cloud, whereas optically-thick lines saturate, only tracing the outer layers. From the peak 13 CO/C18 O ratio,
we can get a measure of the optical depth of the C18 O J = 3 → 2
line in the densest region along the line of sight and thereby test its
fidelity as a total mass tracer.
c 2009 RAS, MNRAS 000, 1–15
9
Using the radiative transfer equation for an isothermal slab
(see e.g. Rohlfs & Wilson 2004), we can relate the 13 CO/C18 O intensity ratio at the velocity of the peak of the C18 O line, R, to the
optical depths of both species (τ13 and τ18 respectively), assuming both transitions are at the same frequency, emanating from the
same volume of material and have the same ηmb and beam-filling
factors:
R=
TA∗ (13 CO) 1 − exp(−τ13 )
.
=
TA∗ (C18 O) 1 − exp(−τ18 )
(2)
The opacities are expected to be linked through the abundance of
13 CO relative to C18 O, which we denote ζ , as τ = ζ τ (e.g. My13
18
ers, Linke & Benson 1983). Where the lines (particularly of 13 CO)
are optically thin, τ → 0, and R → ζ asymptotically. The absolute
values of the 13 CO and C18 O abundances are not known to high
precision and may differ across regions or between clouds, due to
statistical variations across the Galaxy. For instance in photon dominated regions, R can be very high where there is little self-shielding
of the cloud from incident radiation, causing the rarer isotopologue
to be almost completely destroyed (Störzer et al. 2000). Wilson
& Rood (1994) collated various studies together and tracked the
changes in relative abundances with distance from the Galactic centre. In the local interstellar medium they found [12 C]/[13 C] = 77±7
and [16 O]/[18 O] = 560 ± 25, indicating ζ = 7.3 ± 0.7 if there is no
fractionation.
In calculations of R, we only include points that have a good
detection of both the C18 O and 13 CO lines (peak line brightness > 5
times the RMS spectral noise). We plot R in Fig. 11. Where the gas
becomes less dense R tends to ζ . In Fig. 11, outside of the SCUBA
contours, so presumably at lower space densities, the ratio is ∼5–6.
At the very edges of the measured values in NGC1333 and IC348, R
reaches ∼7–8, the value from Wilson & Rood (1994) and becomes
larger still. These larger ratios could either be a result of actual deviations in ζ or from noise effects. Most of the largest R are in
the northeast of NGC1333 and northeast of the main horseshoe of
SCUBA emission in IC348. These two areas are not representative
of the two regions. In NGC1333, this is where a bubble of bright
8.0 µm Spitzer flux resides (see fig. 1 of Gutermuth et al. 2008),
commonly thought to be emission from polycyclic aromatic hydrocarbon features excited by UV radiation. The UV source is likely
to heat up the molecular gas in the region and perhaps change the
isotopologues’ abundances. In the IC348 area, there is a mass of
blue-shifted 12 CO gas, expanding into a dust cavity (Tafalla et al.
2006 and Curtis et al., in preparation). Thus, disregarding these areas, R →∼ 7 at the edges of the map, which we take to be ζ in the
following analysis.
For ζ = 7, τ13 = 7τ18 and R varies with τ as in Fig. 12. R
clearly falls towards the centre of the dust filaments and cores,
where we expect the gas to be more dense. The average R in each
region is unlikely to relate convincingly the properties of the whole
area as it is measured over such a small portion of the maps, particularly in L1448 and L1455. The typical ratios can be measured in
the majority of the star-forming SCUBA cores, which may prove
more informative. We measured R at the peak of each core identified by Hatchell et al. (2007a) and present their averages with those
in the entire regions for comparison in Tables 5 and 6.
The typical ratios across the four regions, R ≈ 2 − 4, translate into τ13 ≈ 1.52 − 0.15 and τ18 ≈ 0.22 − 0.02. This suggests
that the C18 O line is typically optically thin but the 13 CO can be
classed neither as optically thin nor thick. The pattern of R suggests
that on average L1448 has the highest column density, followed by
NGC1333, whilst IC348 has the smallest. The value for L1455 is
10
E. I. Curtis, J. S. Richer and J. V. Buckle
Figure 10. Motions along the line of sight in our target fields, NGC1333 (top), IC348 (bottom left) and L1448 (bottom right). At each point where the C18 O
J = 3 → 2 line has a peak greater than C18 O > 3σRMS the line centre velocity (in km s−1 ) is plotted. The line centres have been computed by fitting a single
Gaussian to the spectrum at each point. L1455 did not have strong enough C18 O emission for such fitting.
Table 5. 13 CO/C18 O ratios, R, at the peak of the SCUBA cores in the
Hatchell et al. (2007a) catalogue. σR is the standard deviation of R across
the cores.
hRi
Region or
Core Class
Number
NGC1333
IC348
L1448
L1455
28
16
7
1
2.8
3.0
2.2
3.6
0.8
0.8
0.3
–
Starless
Class 0
Class I
21
21
10
2.8
2.6
3.1
0.6
0.8
1.0
Table 6. 13 CO/C18 O ratios, R, for every map pixel in Fig. 11 σR is the
standard deviation of R across the pixels.
Region
hRi
σR
Median R
NGC1333
IC348
L1448
L1455
3.7
4.1
2.8
3.6
1.3
1.8
0.7
0.9
3.5
3.6
2.7
3.8
σR
not representative of the region as it is only measured in the dense
cores. When just the cores are examined the pattern is the same
IC348 has the largest R followed by NGC1333 and then L1448.
As we expect, the cores have larger optical depths than their parent
clouds on average. However, the reduction in R is not enough to
render the core C18 O gas optically thick. The smallest core ratio
is 1.62 in NGC1333, which implies τ18 ≈ 1.0 whereas must cores
have much smaller τ18 . The mean core ratios are smaller (and thus
the column densities are higher) for Class 0 than I protostars, although there is significant spread in both populations. As collapse
progresses in protostars, they accrete an increasing fraction of their
envelopes on to the central object so reducing the column density
seen towards Class I over 0 cores. The column density seen towards
starless cores presumably depends on the exact age of the core so
could be comparable to the protostars’ or entirely different.
c 2009 RAS, MNRAS 000, 1–15
11
02
Declination (J2000)
31:56
14
58
16
32:00
18
Declination (J2000)
04
31:20
06
22
08
24
26
The kinematics of the Perseus molecular cloud
45:00
50
40
30
20
10
3:44:00
50
40
43:30
10
12
Right ascension (J2000)
30
20
10
3:29:00
50
40
28:30
Right ascension (J2000)
48
19
ngc1333_13cotoc18o_peak.sdf
3:28:56.715 31:17:57.15 J2000
15
14
Declination (J2000)
44
30:10
30:40
11
41
12
42
13
43
Declination (J2000)
45
16
46
17
47
18
Jun 15, 2009 at 08:54:27
3:26:00
50
40
30
20
10
3:28:00
25:00
50
40
1
2
30
20
10
27:00
Right ascension (J2000)
Right ascension (J2000)
3
4
5
6
7
8
9
10
Figure 11. Ratio of the 13 CO to C18 O antenna temperatures (R) measured at the C18 O line peak in NGC1333 (top left), IC348 (top right), L1448 (bottom left)
and L1455 (bottom right). Pixels are only displayed where both isotopologues have peak intensities > 5σRMS . Contours as in Fig. 10.
5.3
13 CO/C18 O
Temperature
12 CO
In this section, we derive Tex for the
J = 3 → 2 line assuming it is optically thick. We include 12 CO HARP data for NGC1333
over the region observed in 13 CO/C18 O (J. Swift, personal communication) which are distributed on a coarser grid of 6 arcsec pixels
using a nearest-neighbour scheme.
In LTE, Tex is the physical temperature of the gas, although
this will of course not be the case in general. We have noted typical 13 CO optical depths of τ13 ≈ 1 − 2, which provided the abundance [12 C]/[13 C] = 77 (Wilson & Rood 1994) should ensure the
opacity of the 12 CO gas is τ12 ≈ 77 − 154. Ratio maps with other
isotopologues for 12 CO are unlikely to produce accurate τ12 values as 12 CO traces outer cloud layers and not the volumes where
c 2009 RAS, MNRAS 000, 1–15
are emitted. However, provided the line is optically
thick, i.e. τ → ∞, and not self-absorbed, we can derive Tex thus
(see e.g. Pineda et al. 2008):
Tex (12 CO) =
16.59 K
ln 1 + 16.6 K/[Tmax (12 CO) + 0.036 K]
(3)
where T0 = 16.59 K and Tmax (12 CO) is the peak 12 CO main beam
brightness temperature. We plot Tex in Figs. 13 and 14. Tex is typically 5–25 K, generally increasing towards the centre of each region
where it can reach 30 K.
There are a number of caveats to this analysis that should be
noted. First, in some respects the most interesting regions of molecular clouds are the cores in which stars are formed. The tempera-
12
E. I. Curtis, J. S. Richer and J. V. Buckle
10
20
15
25
30
35
40
45
24
26
5
10
15
20
25
30
35
04
02
Declination (J2000)
32:00
18
31:56
14
58
16
Declination (J2000)
31:20
06
22
08
5
45:00
50
40
30
20
10
3:44:00
50
40
43:30
10
12
Right ascension (J2000)
30
20
10
3:29:00
50
40
28:30
Right ascension (J2000)
4
6
8
10
12
14
16
18
10
5
20
15
20
25
15
13
14
Declination (J2000)
45
44
43
3:26:00
50
40
30
20
10
25:00
Right ascension (J2000)
09
30:40
30:10
41
11
12
42
Declination (J2000)
16
46
17
47
18
48
19
2
3:28:00
50
40
30
20
10
27:00
Right ascension (J2000)
Figure 13. Tex in K derived from the peak 12 CO main beam brightness temperature with panels as arrange in Fig. 11. Contours as in Fig. 10.
tures of such high-density regions will not be accurately probed by
12 CO, whose emission is restricted to outer cloud material due to its
higher opacity. However, we might naturally expect the temperature
of starless cores (without an internal heating source) and possibly
protostars to reflect that of their natal environment so the 12 CO temperature should reflect the pattern of core conditions. Second, 12 CO
may not be optically thick everywhere, although the typical 13 CO
optical depths in regions with bright SCUBA emission do suggest
it will be for most of the areas. Furthermore, we may underestimate
the true 12 CO peak temperature if the line profile is complicated by
effects such as infall and self-absorption.
In L1455, L1448 and IC348, the Tex maps follow the integrated 12 CO intensities closely. Any strong outflows are promi-
nent areas of higher Tex , as suggested by other authors (Hatchell,
Fuller & Ladd 1999; Nisini et al. 2000). The dominant feature in
NGC1333’s Tex map is the blue-shifted bubble of gas in the northeast corner. We have already mentioned this is an area likely to
be illuminated by a UV source. By contrast, the most pronounced
feature in the 12 CO integrated intensity, the bright outflow from
SVS13, is almost absent. The bubble’s Tex is some 40 K higher than
the maximum in the other regions. In this part of the map, 12 CO is
not optically thick and the gas probably has a higher physical temperature – as we noted the 13 CO/C18 O ratio is larger in this region.
Additionally, it only has single-peaked lines, devoid of some of the
more complicated profiles seen in NGC1333, suggesting its peak
temperature is over-estimated compared to the rest of the cloud.
c 2009 RAS, MNRAS 000, 1–15
The kinematics of the Perseus molecular cloud
Table 7. Tex from 12 CO at the peak of the SCUBA cores in the Hatchell
et al. (2007a) catalogue by region or source age.
Number
hTex i/K
σT /K
NGC1333
IC348
L1448
L1455
29
17
7
5
25.2
17.6
10.6
12.6
10.1
3.2
2.0
2.2
Starless
Class 0
Class I
23
21
14
18.8
18.3
24.9
9.7
6.2
11.2
Region or Core Class
This all points to a warm, low-density region with material perhaps dispersed by nearby luminous stars, which Hatchell, Fuller &
Richer (2007b) suggest are the cause of a lack of outflow detections
in this area.
It is interesting that the largest temperatures in
NGC1333 are to the northeast, in the direction of 40 Persei
(03h 42m 22.6s ,+33◦ 570 54.100 ), a B0.5 star, part of the Perseus OB
association, suggested to be triggering the star formation in these
clouds (Walawender et al. 2004; Kirk et al. 2006). Indeed, the
highest temperatures are to the north in IC348, also in the direction
of 40 Per, although no such gradients are clear in L1448 and
L1455, possibly as they are further away.
On average, L1448 is the coldest, followed in order of increasing temperature by L1455 then IC348, with NGC1333 the hottest.
The histograms of Fig. 14 show two distinct temperature environments: one colder in L1455 and L1448 and the other hotter in IC348
and NGC1333. The latter two regions show a large range of temperatures. NGC1333 and IC348 are probably heated more by their
namesake IR clusters and the triggering radiation from 40 Per than
L1448 and L1455. Protostars may also heat their surroundings to a
limited extent via radiation and outflows. NGC1333 with its numerous energetic flows and protostars is most likely to be warmed up
in this way. L1448 has a high protostellar fraction as well and powerful outflows but less ambient gas with which to interact – most of
the outflows have broken out of the region.
As we have already mentioned, temperatures derived from
12 CO are unlikely to be best probes of conditions inside dense
cores. Nevertheless we might expect the core temperatures to be
affected to some extent by the temperature in the bulk of the gas.
From Table 7, it is clear the temperatures towards the cores are
larger on average than within their parent region as a whole but
exhibit the same trends. When comparing cores of different ages
it is worth noting that we expect fundamentally different relations
between Tex and the true temperatures for starless and protostellar
cores, Tcore . Towards the centres of starless cores, temperatures fall
whereas for protostars temperatures rise. Thus, for starless cores
Tex > Tcore and for protostars Tex < Tcore . In Table 7 the environments of Class I protostars appear slightly warmer than both starless cores and Class 0 protostars on average. As a source ages we
expect its core temperature to increase, resulting from the increase
in brightness of its central object. These data would exhibit this
trend except the Class 0 and starless cores have similar temperatures, probably illustrating that the 12 CO line probes an environment outside the high-density core, where Tex > Tcore for the starless cores and Tex < Tcore for the protostars.
We can make an interesting comparison between this gas temperature and that of the dust. At high densities, n & 2 × 104 cm−3 ,
conditions likely to be probed in CO with HARP, the gas and dust
c 2009 RAS, MNRAS 000, 1–15
13
temperatures are coupled (Galli, Walmsley & Gonçalves 2002).
Recently, Schnee et al. (2008) computed the dust temperature at
40 arcsec resolution across Perseus using Spitzer and IRAS measurements. We plot Tex degraded to 40 arcsec resolution versus
Schnee et al.’s dust temperature in Fig. 15. The dust temperature,
TD , was calculated from Spitzer 70 and 160 µm fluxes, which is
dominated by warm dust along the line of sight rather than the
denser, colder material in say starless cores and is more likely to
match the material traced in 12 CO. There is actually little correlation between the two temperatures. Typically TD is slightly lower
than Tex : 12–20 K compared to 8–30 K. In NGC1333 and IC348,
Tex is mostly larger than TD although there is a significant proportion of pixels where the opposite is true. However, in L1448 and
L1455, nearly everywhere TD > Tex . This may be explained if the
12 CO gas is not tracing the same areas as the dust emission, with
the 12 CO emitted in a colder layer whilst the dust comes from a hotter inner cloud region. Additionally, if the gas is not optically thick,
then we would over-estimate the 12 CO excitation temperature and
Tex > TD .
The 12 CO excitation temperature may be correlated with the
cloud column density. As the rotational transition is collisionally
excited, on moving to denser and higher extinction portions of the
cloud with densities above the critical density, the gas Tex should increase. We use the integrated C18 O J = 3 → 2 intensity as a proxy
for the visual extinction since typically there is a linear relation between them. Using C18 O J = 1 → 0 data across the Perseus molecular cloud Pineda et al. (2008) measured:
R
Tmb (C18 O)dv
+ (2.9 ± 0.9).
(4)
Av /mag = (2.4 ± 0.1)
K km s−1
In Fig. 15, we plot the C18 O integrated intensity, TA∗ dv versus
Tex , with lines denoting the expected variation for various column
densities of C18 O assuming LTE. No single line is a good fit for
all the points as in each region there is a range of column densities.
However, a moderate range of column densities will span the whole
parameter space. At low temperatures with a good deal of scatter
the integrated intensity increases linearly with Tex . L1448 again requires the highest column densities with IC348 and NGC1333 having a large spread in conditions.
R
6
SUMMARY
This paper presents the technical details and preliminary analysis of a large-scale survey of the kinematics of molecular gas in
the Perseus molecular cloud. Observations of the J = 3 → 2 rotational transitions of 12 CO, 13 CO and C18 O in over 600 arcmin2 of
NGC1333, IC348, L1448 and L1455 were undertaken with HARP
on the JCMT. We introduce a new ‘flatfield’ procedure to account
for striping artefacts in HARP scan maps of molecular clouds, apparently resulting from differential performance across the detectors and/or their samplers. The data from each working detector
is multiplied by a conversion factor which scales its intensity to
match the nominated reference detector. The factors are computed
from the scientific observations themselves by calculating the total
intensity received by each detector across the whole map.
We compare integrated intensity maps of the three tracers to
SCUBA 850 µm emission (Hatchell et al. 2005) and the position of
protostars and starless cores (Hatchell et al. 2007a) in each field.
There is a striking similarity between the SCUBA maps and C18 O
emission, hinting that similar densities of material are traced with
the gas and dust strongly coupled. However, the detailed structure
14
E. I. Curtis, J. S. Richer and J. V. Buckle
of individual star-forming cores is not always so simple, for example very bright SCUBA cores sometimes have only weak C18 O
emission and vice versa. Many outflows are obvious in the 12 CO
data, which we will examine in detail in a subsequent study (Curtis
et al. in preparation). The 13 CO maps are somewhat intermediate
between the other two isotopologues and average spectra across the
maps do show weak linewings from outflows. Intriguingly, masses
derived from the HARP and SCUBA data exhibit different trends
across the four regions, emphasizing the need to examine the detailed excitation conditions across regions rather than simple constant assumptions.
From the 13 CO/C18 O integrated line ratio, R, we explore variations in the gas opacity and estimate the two species’ relative
abundance ([13 CO]/[C18 O] ∼ 7). Across most of the regions the
C18 O gas is optically thin (τ18 = 0.02 − 0.22) and therefore is a reliable total mass tracer. Indeed, inside the denser parts of the clouds,
the star-forming cores, the opacity does not increase beyond a maximum, τ18 = 0.9. When we also consider that the critical density of
the transition is some 104 cm−3 , we expect to probe an intermediate region between a dense core and its envelope with the C18 O
line. The 13 CO optical depths are neither in the optically thin nor
thick regimes (τ13 is typically 0.15–1.52). Class 0 protostars have
smaller ratios than Class Is as expected as more material accretes
on to the central object over time.
If we assume the 12 CO line is optically thick, an estimate of
the excitation temperature can be gathered from the peak line temperature, assuming LTE (in which case Tex is the physical gas temperature). In general we derive temperature of 5–25 K, increasing
towards the centres of the individual regions and in outflow lobes.
An area in the northwest of NGC1333 has temperatures of over
45 K, probably as it is heated by a UV source and/or the region has
lower density so is not optically thick. IC348 and NGC1333 are on
average much warmer than L1448 and L1455. This is partly as the
averages for NGC1333 are skewed upwards because of the warm
gas in the north and perhaps since NGC1333 and IC348 are closer
to 40 Per (thought to be triggering star formation in Perseus). There
is little correlation between Tex and the dust temperature, TD , derived from Spitzer observations (Schnee et al. 2008). Typically TD
is slightly lower than Tex : 12–20 K compared to 8–30 K. However,
in L1448 and L1445, typically Tex < TD , this may imply we are
tracing completely different regions with the 12 CO gas and warm
dust or that the gas is not optically thick, so we are over-estimating
Tex .
This work demonstrates the utility of HARP for large-scale
surveys of gas kinematics in nearby molecular clouds. The J = 3 →
2 transitions of CO and its isotopologues are powerful probes of the
conditions of star formation when used in combination, examining
moderately high densities comparable to the dust densities seen by
SCUBA.
7
ACKNOWLEDGMENTS
EIC thanks the Science and Technology Facilities Council (STFC)
for studentship support while carrying out this work. The authors
thank Jonathan Swift for use of the 12 CO data towards NGC1333 in
advance of publication. We are also grateful to the referee, whose
useful comments and suggestions significantly improved the clarity of this paper. The JCMT is operated by The Joint Astronomy
Centre (JAC) on behalf of the STFC of the United Kingdom, the
Netherlands Organisation for Scientific Research and the National
Research Council (NRC) of Canada. We have also made exten-
sive use of the SIMBAD data base, operated at CDS, Strasbourg,
France. We acknowledge the data analysis facilities provided by
the Starlink Project which is maintained by JAC with support from
STFC. This research used the facilities of the Canadian Astronomy
Data Centre operated by the NRC with the support of the Canadian
Space Agency.
REFERENCES
Bachiller R., Cernicharo J., 1986a, A&A, 166, 283
Bachiller R., Cernicharo J., 1986b, A&A, 168, 262
Bachiller R., Martin-Pintado J., Tafalla M., Cernicharo J.,
Lazareff B., 1990, A&A, 231, 174
Bachiller R., Guilloteau S., Dutrey A., Planesas P., Martin-Pintado
J., 1995, A&A, 299, 857
Ballesteros-Paredes J., Mac Low M.-M., 2002, ApJ, 570, 734
Bally J., Lada E. A., Lane A. P., 1993, ApJ, 418, 322
Bally J., Walawender J., Johnstone D., Kirk H., Goodman A.,
2008, in Reipurth B., ed., ASP Conf. Ser., Handbook of Star
Forming Regions. Astron. Soc. Pac., San Francisco, p. 308
Bate M. R., 2009a, MNRAS, 392, 590
Bate M. R., 2009b, MNRAS, 392, 1363
Bate M. R., Bonnell I. A., 2005, MNRAS, 356, 1201
Blake G. A., Sandell G., van Dishoeck E. F., Groesbeck T. D.,
Mundy L. G., Aspin C., 1995, ApJ, 441, 689
Boulard M.-H., Caux E., Monin J.-L., Nadeau D., Rowlands N.,
1995, A&A, 300, 276
Buckle J. V., Hills, R. E., Smith, H. et al., 2009, MNRAS, in press
Černis K., 1990, Astrophys. and Space Sci., 166, 315
Černis K., 1993, Baltic Astron., 2, 214
Černis K., Straižys V., 2003, Baltic Astron., 12, 301
Chandler C. J., Richer J. S., 2001, ApJ, 555, 139
Cohen M., Jones B. F., Hereld M., 1991, ApJ, 371, 237
Curiel S., Raymond J. C., Moran J. M., Rodriguez L. F., Canto J.,
1990, ApJ, 365, L85
Curiel S., Torrelles J. M., Rodrı́guez L. F., Gómez J. F., Anglada
G., 1999, ApJ, 527, 310
Davis C. J., Eislöffel J., Ray T. P., Jenness T., 1997, A&A, 324,
1013
Davis C. J., Scholz P., Lucas P., Smith M. D., Adamson A., 2008,
MNRAS, 387, 954
de Zeeuw P. T., Hoogerwerf R., de Bruijne J. H. J., Brown
A. G. A., Blaauw A., 1999, AJ, 117, 354
Di Francesco J., Johnstone D., Kirk H., MacKenzie T., Ledwosinska E., 2008, ApJS, 175, 277
Eislöffel J., Froebrich D., Stanke T., McCaughrean M. J., 2003,
ApJ, 595, 259
Enoch M. L.,Young K. E., Glenn J. et al., 2006, ApJ, 638, 293
Enoch M. L., Evans II N. J., Sargent A. I., Glenn J., Rosolowsky
E., Myers P., 2008, ApJ, 684, 1240
Enoch M. L., Evans II N. J., Sargent A. I., Glenn J., 2009, ApJ,
692, 973
Evans II N. J., Allen L. E., Blake G. A. et al., 2003, PASP, 115,
965
Evans II N. J., Dunham M. M., Jørgensen J. K. et al., 2009, ApJS,
181, 321
Galli D., Walmsley M., Gonçalves J., 2002, A&A, 394, 275
Goldsmith P. F., Snell R. L., Hemeon-Heyer M., Langer W. D.,
1984, ApJ, 286, 599
Grossman E. N., Masson C. R., Sargent A. I., Scoville N. Z., Scott
S., Woody D. P., 1987, ApJ, 320, 356
c 2009 RAS, MNRAS 000, 1–15
The kinematics of the Perseus molecular cloud
Gueth F., Guilloteau S., 1999, A&A, 343, 571
Gutermuth R. A., Myers P. C., Megeath S. T. et al., 2008, ApJ,
674, 336
Haschick A. D., Moran J. M., Rodriguez L. F., Burke B. F., Greenfield P., Garcia-Barreto J. A., 1980, ApJ, 237, 26
Hatchell J., Fuller G. A., Ladd E. F., 1999, A&A, 344, 687
Hatchell J., Richer J. S., Fuller G. A., Qualtrough C. J., Ladd E. F.,
Chandler C. J., 2005, A&A, 440, 151
Hatchell J., Fuller G. A., Richer J. S., Harries T. J., Ladd E. F.,
2007a, A&A, 468, 1009
Hatchell J., Fuller G. A., Richer J. S., 2007b, A&A, 472, 187
Herbig G. H., 1974, Lick Obser. Bull., 658, 1
Herbig G. H., Jones B. F., 1983, AJ, 88, 1040
Herbst W., 2008, in Reipurth B., ed., ASP Conf. Ser., Handbook
of Star Forming Regions. Astron. Soc. Pac., San Francisco, p.
372
Hirota T., Bushimata T., Choi Y. K. et al., 2008, PASJ, 60, 37
Ho P. T. P., Barrett A. H., 1980, ApJ, 237, 38
Holland W. S., Robson E. I., Gear W. K. et al., 1999, MNRAS,
303, 659
Jenness T., Cavanagh B., Economou F., Berry D. S., 2008, in Argyle R. W., Bunclark P. S., Lewis J. R., eds, ASP Conf. Ser., Vol.
394, Astronomical Data Analysis Software and Systems XVII.
Astron. Soc. Pac., San Francisco, p. 565
Jennings R. E., Cameron D. H. M., Cudlip W., Hirst C. J., 1987,
MNRAS, 226, 461
Johnstone D., Wilson C. D., Moriarty-Schieven G., Joncas G.,
Smith G., Gregersen E., Fich M., 2000, ApJ, 545, 327
Jørgensen J. K., Harvey P. M., Evans II N. J. et al., 2006, ApJ,
645, 1246
Jørgensen J. K., Johnstone D., Kirk H., Myers P. C., 2007a, ApJ,
656, 293
Jørgensen J. K., Bourke T. L., Myers P. C. et al., 2007b, ApJ, 659,
479
Jørgensen J. K., Johnstone D., Kirk H., Myers P. C., Allen L. E.,
Shirley Y. L., 2008, ApJ, 683, 822
Kirk H., Johnstone D., Di Francesco J., 2006, ApJ, 646, 1009
Kirk H., Johnstone D., Tafalla M., 2007, ApJ, 668, 1042
Klessen R. S., Burkert A., 2000, ApJS, 128, 287
Knee L. B. G., Sandell G., 2000, A&A, 361, 671
Kutner M. L., Ulich B. L., 1981, ApJ, 250, 341
Lada C. J., Alves J., Lada E. A., 1996, AJ, 111, 1964
Ladd E. F., Lada E. A., Myers P. C., 1993, ApJ, 410, 168
Lay O. P., Carlstrom J. E., Hills R. E., 1995, ApJ, 452, L73
Lefloch B., Castets A., Cernicharo J., Langer W. D., Zylka R.,
1998, A&A, 334, 269
Levreault R. M., 1988, ApJS, 67, 283
Lightfoot J. F., Glencross W. M., 1986, MNRAS, 221, 993
Looney L. W., Mundy L. G., Welch W. J., 2000, ApJ, 529, 477
Luhman K. L., Rieke G. H., Lada C. J., Lada E. A., 1998, ApJ,
508, 347
Luhman K. L., Stauffer J. R., Muench A. A., Rieke G. H., Lada
E. A., Bouvier J., Lada C. J., 2003, ApJ, 593, 1093
McCaughrean M. J., Rayner J. T., Zinnecker H., 1994, ApJ, 436,
L189
Minchin N. R., White G. J., Padman R., 1993, A&A, 277, 595
Motte F., Andre P., Neri R., 1998, A&A, 336, 150
Myers P. C., Linke R. A., Benson P. J., 1983, ApJ, 264, 517
Nisini B., Benedettini M., Giannini T., Codella C., Lorenzetti D.,
di Giorgio A. M., Richer J. S., 2000, A&A, 360, 297
Olmi L., Testi L., 2002, A&A, 392, 1053
c 2009 RAS, MNRAS 000, 1–15
15
Padoan P., Juvela M., Goodman A. A., Nordlund Å., 2001, ApJ,
553, 227
Pineda J. E., Caselli P., Goodman A. A., 2008, ApJ, 679, 481
Price D. J., Bate M. R., 2008, MNRAS, 385, 1820
Quillen A. C., Thorndike S. L., Cunningham A., Frank A., Gutermuth R. A., Blackman E. G., Pipher J. L., Ridge N., 2005, ApJ,
632, 941
Rebull L. M., Stapelfeldt K. R., Evans II N. J. et al., 2007, ApJS,
171, 447
Ridge N. A., Schnee S. L., Goodman A. A., Foster J. B., 2006a,
ApJ, 643, 932
Ridge N. A., Di Francesco J., Kirk H. et al., 2006b, AJ, 131, 2921
Rodrı́guez L. F., Anglada G., Curiel S., 1999, ApJS, 125, 427
Rohlfs K., Wilson T. L., 2004, Tools of Radio Astronomy.
Springer, Berlin
Sandell G., Knee L. B. G., 2001, ApJ, 546, L49
Schnee S., Li J., Goodman A. A., Sargent A. I., 2008, ApJ, 684,
1228
Schöier F. L., van der Tak F. F. S., van Dishoeck E. F., Black J. H.,
2005, A&A, 432, 369
Smith H., Hills R. E., Withington S. et al., 2003, in Phillips T. G.,
Zmuidzinas J., eds, Proc. SPIE Vol. 4855, Millimeter and Submillimeter Detectors for Astronomy. SPIE, Bellingham, p. 338
Smith H., Buckle J., Hills R. et al., 2008a, in Duncan W. D., Holland W. S., Withington S., Zmuidzinas J., eds, Proc. SPIE, Vol.
7020, Millimeter and Submillimeter Detectors and Instrumentation for Astronomy IV. SPIE, Bellingham, p. 24
Smith R. J., Clark P. C., Bonnell I. A., 2008b, MNRAS, 391, 1091
Snell R. L., Edwards S., 1981, ApJ, 251, 103
Störzer H., Zielinsky M., Stutzki J., Sternberg A., 2000, A&A,
358, 682
Strom S. E., Vrba F. J., Strom K. M., 1976, AJ, 81, 314
Tafalla M., Kumar M. S. N., Bachiller R., 2006, A&A, 456, 179
Visser A. E., Richer J. S., Chandler C. J., 2001, MNRAS, 323,
257
Walawender J., Bally J., Reipurth B., Aspin C., 2004, AJ, 127,
2809
Walawender J., Bally J., Kirk H., Johnstone D., Reipurth B., Aspin
C., 2006, AJ, 132, 467
Walawender J., Bally J., Di Francesco J., Jørgensen J., Getman
K., 2008, in Reipurth B., ed., ASP Conf. Ser., Handbook of Star
Forming Regions. Astron. Soc. Pac., San Francisco, p. 346
Walsh A. J., Myers P. C., Di Francesco J., Mohanty S., Bourke
T. L., Gutermuth R., Wilner D., 2007, ApJ, 655, 958
Ward-Thompson D., Di Francesco J., Hatchell J. et al., 2007,
PASP, 119, 855
Warin S., Castets A., Langer W. D., Wilson R. W., Pagani L.,
1996, A&A, 306, 935
Wilking B. A., Meyer M. R., Greene T. P., Mikhail A., Carlson
G., 2004, AJ, 127, 1131
Wilson T. L., Rood R., 1994, ARA&A, 32, 191
Wolf-Chase G. A., Barsony M., O’Linger J., 2000, AJ, 120, 1467
This paper has been typeset from a TEX/ LATEX file prepared by the
author.
16
E. I. Curtis, J. S. Richer and J. V. Buckle
Bolo114
600
400
800
1000
1200
1600
1400
0.5
1
1.5
2
2.5
3
3.5
4
24
26
04
15
13
23
Declination (J2000)
02
21
16
14
101
12
19
32:00
25
02
17
18
20
32:00
Declination (J2000)
04
06
06
08
08
200
Bolo113
50
40
30
20
10
3:44:00
50
40
43:30
Right ascension (J2000)
1
1.5
45:00
2
2.5
3
40
10
30
20
30
10
40
3:44:00
50
40
60
50
43:30
70
08
20
06
06
02
Declination (J2000)
04
04
02
Declination (J2000)
32:00
58
58
31:56
31:56
45:00
50
Right ascension (J2000)
3.5
08
0.5
32:00
45:00
31:56
31:56
58
58
Bolo111
50
40
30
20
10
3:44:00
50
40
43:30
Right ascension (J2000)
4
6
8
10
50
40
30
20
10
3:44:00
50
40
43:30
Right ascension (J2000)
12
14
16
18
02
32:00
31:56
58
Declination (J2000)
04
06
08
2
45:00
45:00
50
40
30
20
10
3:44:00
50
40
43:30
Right ascension (J2000)
Figure 6. Overview of the data towards IC348 as for NGC1333 in Fig. 5.
Top: SCUBA 850 µm emission. Prominent sources in the field are labelled:
HH211 (McCaughrean et al. 1994; Gueth & Guilloteau 1999; Chandler
& Richer 2001), IC348-MMS (Eislöffel et al. 2003; Tafalla et al. 2006;
Walawender et al. 2006) and IC348-SMM3 (Tafalla et al. 2006). Middle: Integrated C18 O J = 3 → 2 intensity from 7 to 10 km s−1 with contours from 0.5 to 3.0 K km s−1 in 0.5 K km s−1 steps. Bottom: Integrated
13 CO J = 3 → 2 intensity from 5 to 12 km s−1 with contours from 2 to
18 K km s−1 in 2 K km s−1 steps.
Figure 6 – continued Top: Contours of SCUBA 850 µm emission (levels
as previously) overlaid on the C18 O J = 3 → 2 integrated intensity as before. Middle: Integrated 12 CO intensity from 0 to 20 km s−1 with contours
from 15 to 75 K km s−1 in steps of 5 K km s−1 . Bottom: Red-green-blue
colour composite image of the mean 12 CO data value in various velocity
ranges: from 3.5 to 5.5 km s−1 (blue), 7.0 to 9.0 km s−1 (green) and 12.0 to
14.0 km s−1 (red).
c 2009 RAS, MNRAS 000, 1–15
17
The kinematics of the Perseus molecular cloud
1000
1500
2000
2500
3000
0.5
1
1.5
2
2.5
3
3.5
4
46
46
47
47
48
48
500
27
30
Declination (J2000)
29
30:40
30:40
41
41
42
42
32
43
44
Bolo11
44
45
31
43
Declination (J2000)
45
28
50
3:26:00
40
30
20
10
25:00
Right ascension (J2000)
3:26:00
50
40
30
20
10
25:00
Right ascension (J2000)
1
1.5
2
2.5
3
3.5
4
10
30
40
50
60
70
80
90
47
47
44
30:40
30:40
41
41
42
43
Declination (J2000)
45
46
46
45
44
43
42
Declination (J2000)
20
48
48
0.5
50
3:26:00
40
30
20
10
25:00
Right ascension (J2000)
2
6
4
3:26:00
50
40
30
20
10
25:00
Right ascension (J2000)
8
10
12
Starlink GAIA::Skycat
L1448
l1448_co_low_col.sdf
3:25:30.427 30:43:56.51 J2000
Jan 09, 2009 at 16:21:54
45
44
43
30:40
41
42
Declination (J2000)
46
47
48
Meredith
3:26:00
50
40
30
20
10
25:00
Right ascension (J2000)
Figure 7. Overview of the data towards L1448 as for IC348 in Fig. 6.
Top: SCUBA 850 µm emission. Prominent sources in the field are labelled:
L1448NW, L1448:A/B (Curiel et al. 1990, 1999), L1448C (Bachiller et al.
1990, 1995) and L1448-IRS2 (Wolf-Chase et al. 2000).
c 2009 RAS, MNRAS 000, 1–15
Figure 7 – continued Top: Contours of SCUBA 850 µm emission (levels as
previously) overlaid on the C18 O J = 3 → 2 integrated intensity as before.
Middle: Integrated 12 CO intensity from −25 to 25 km s−1 with contours
from 10 to 90 K km s−1 in steps of 10 K km s−1 . Bottom: Ranges: from −25
to 0 km s−1 (blue), 2 to 6 km s−1 (green) and 7 to 14 km s−1 (red).
18
E. I. Curtis, J. S. Richer and J. V. Buckle
200
300
400
600
500
700
0.3
0.5
0.6
0.7
0.8
0.9
1
1.1
18
18
17
17
15
40
12
12
36
13
35
14
Declination (J2000)
16
16
15
14
39
13
Declination (J2000)
0.4
19
19
100
30:10
30:10
11
11
37
3:28:00
50
40
30
20
10
27:00
3:28:00
50
40
Right ascension (J2000)
0.6
0.4
0.8
1
1.2
20
30
20
40
10
50
27:00
60
1.4
3:28:00
50
40
30
20
10
27:00
Right ascension (J2000)
15
09
30:10
30:10
11
11
12
12
13
13
14
Declination (J2000)
15
14
Declination (J2000)
16
16
17
17
18
18
19
19
0.2
30
Right ascension (J2000)
10
3:28:00
50
40
30
20
10
27:00
Right ascension (J2000)
2
3
4
5
6
10
27:00
15
14
09
30:10
11
12
13
Declination (J2000)
16
17
18
19
1
3:28:00
50
40
30
20
Right ascension (J2000)
Figure 8. Overview of the data towards L1455 as for IC348 in Fig. 6. Top:
SCUBA 850 µm emission. Additionally labelled is L1455-FIR (Davis et al.
1997).
Figure 8 – continued Top: Contours of SCUBA 850 µm emission (levels as
previously) overlaid on the C18 O J = 3 → 2 integrated intensity as before.
Middle: Integrated 12 CO intensity from −5 to 15 km s−1 with contours from
15 to 65 K km s−1 in steps of 5 K km s−1 . Bottom: Ranges: from −5.0 to
3.5 km s−1 (blue), 5.0 to 6.5 km s−1 (green) and 8.0 to 14.0 km s−1 (red).
c 2009 RAS, MNRAS 000, 1–15
The kinematics of the Perseus molecular cloud
NGC1333
IC348
L1448
L1455
0.3
Fraction of pixels
19
0.2
0.1
0
5
10
15
20
25
30
35
Tex /K
Region
Figure 9. Average cloud spectra for the different regions. Each plot contains the averages of every spectrum in the HARP J = 3 → 2 cubes: 12 CO
(dashed), 13 CO (dot-dashed) and C18 O (solid).
CO/C18 O line ratio, R
σT (K)
Median Tex (K)
19.3
15.5
8.2
9.3
7.0
3.6
1.0
1.6
17.7
15.0
8.0
9.2
Figure 14. Tex derived from peak 12 CO values. σT is the standard deviation
of Tex .
7
13
NGC1333
IC348
L1448
L1455
hTex i (K)
6
13
CO
C18 O
5
4
3
2
1
0
0
2
4
6
8
τ13 or τ18
10
12
Figure 12. Variation of R with τ for ζ = 7.
c 2009 RAS, MNRAS 000, 1–15
14
20
E. I. Curtis, J. S. Richer and J. V. Buckle
12
CO gas temperature/K
50
40
30
20
10
10
15
20
25
30
Dust temperature/K
8 × 1015 cm−2
C18 O integrated intensity/K km s−1
10
4 × 1015 cm−2
5
2 × 1015 cm−2
1015 cm−2
0
10
30
12
50
70
CO gas temperature/K
Figure 15. Points mark pixels in the various regions’ maps: NGC1333
(blue), IC348 (red), L1448 (green) and L1455 (orange). Top: Tex derived
from the peak 12 CO TA∗ versus the dust temperature derived from Spitzer
MIPS data at 70 and 160 µm with 40 arcsec resolution (Schnee et al. 2008).
The line marks where both temperatures are equal. Bottom: C18 O integrated
intensity versus 12 CO derived Tex . The lines denoted the anticipated dependence for constant column densities of C18 O as labelled.
c 2009 RAS, MNRAS 000, 1–15